摘要
针对传统短时傅里叶变换(STFT)对时变显著信号分辨率不高的缺点,提出了一种对窗函数加调频系数的改进算法,用VC程序予以实现;并对一组三峡工地实测相控阵探地雷达(PGPR)回波数据进行处理,依据不同观测目的,对处理结果使用测点切片视图和频率切片视图进行显示。通过对视图的比较分析,结果表明在改进算法中选择合适的窗函数和窗参数,能有效提高时变显著信号的分辨率,对地质情况进行较精确的定位。
According to the low resolution of significantly time-varying signal analyzed by conventional STFT(Short Time Fourier Transform), this paper introduces an improved algorithm that considers coefficient of frequency modulation into window function. The algorithm is realized by VC++ and applied to a set of Phased-array Ground Penetrating Radar (PGPR) data collected in Sanxia. The processing result is displayed in point section graph and frequency section graph according to various physical meanings. By comparing and analyzing the section graphs, it finds that choosing suitable window functions and para￣meters can effectively improve the resolution of significantly time-varying signal and locate geological objects more precisely.
出处
《计算机应用研究》
CSCD
北大核心
2005年第6期127-129,共3页
Application Research of Computers
基金
国家"863"计划资助项目(2001AA13205003)
国家自然科学基金资助项目(50099620)
长江水利委员会联合资助项目
